The actual task of supplementing intelligent systems of evidence-based medicine with technically implemented mental
models is set. Using these models, the user understands the results of digital models in Big Data systems. Clarified concepts related to this issue. The electronic components of neural networks for the implementation of mental models is defined. A variant of continuous logic of mental models is proposed. Functional expressions of convolutions of input signals of artificial neurons are constructed. The basic operations for use in the computational architecture of a neural network are defined. Prospects for the development of this issue are outlined.